Final Progress Report
Proposal No. IBD-0031R
Principal Investigator: David S. Rampton, BM, DPhil
Applicant Organization: Queen Mary, University of London (United Kingdom)
Project Title: Stress and hypnosis in ulcerative colitis
Period of Award: August 1, 2003 - July 31, 2005
1. Summary of project aims
Psychological stress is often reported as an exacerbating factor in inflammatory bowel disease (IBD). It is thought to act through as yet incompletely defined neuroimmunomodulatory pathways. Hypnotherapy relates to a formalized treatment approach using the induction of trance to treat specific problems with relaxation, suggestion and imagery. It has been shown to be an effective treatment in irritable bowel syndrome, non-ulcer dyspepsia and relapsing duodenal ulcer disease.
The aim of this study was to investigate the hypothesis that inflammation in IBD could be modified by stress levels. We proposed that increased psychological stress worsened inflammation and that relaxation achieved by hypnosis reduced inflammation. We investigated the contrasting pro- and anti -inflammatory effects of psychological stress (administered as an IQ test with distracting auditory input) and hypnosis on the systemic and rectal mucosal components of the inflammatory response in patients with UC and healthy controls. Before and after these procedures, the level of systemic inflammation was assessed by measurement of serum cytokine concentrations and cytokine production by LPS-stimulated whole blood, leucocyte count, Natural Killer (NK) cell numbers, platelet activation and platelet-leucocyte aggregate formation. Rectal mucosal function was evaluated by measurement of the release of TNF-alpha and IL-13, substance P, and mast cell mediators using an in vivo filter paper technique, of production of reactive oxygen metabolites from biopsies in vitro and by the recording of blood flow in vivo using laser doppler flowmetry. The autonomic response to stress was quantified by measurement of pulse and blood pressure. The relationship between the response to stress and hypnotherapy, and chronic stress and anxiety was measured using various well-validated psychometric tests.
2. Accomplishments towards meeting those aims
We were able to achieve the majority of the aims stated in the original grant proposal during the two year funded period.
A stress protocol was developed which was successful in inducing a subjective and objective autonomic stress response in the majority of subjects. A hypnotherapy protocol was devised which was successful in inducing trance in 16 of the 17 subjects recruited.
The required assays to assess the response to stress and hypnotherapy were developed and validated.
Twenty-five patients with inactive UC were recruited to undergo the stress protocol and 17 patients with active UC to undergo the hypnotherapy protocol. Ten patients with inactive UC and eight patients with active UC underwent the control protocol. Ten healthy volunteers underwent the stress protocol, 10 healthy volunteers underwent the hypnotherapy protocol and 10 healthy volunteers underwent the control protocol. The numbers of subjects recruited was greater than that stated in the original grant proposal. We believe that this has adequately allowed us to assess the inflammatory response to stress and hypnotherapy in patients with UC and to make a comparison to healthy volunteers.
One aim of the original project which we were not able to achieve was to assess any anti-inflammatory response to a longer course of hypnotherapy. We were able to recruit only two patients with chronic active UC to undergo six weekly sessions of hypnotherapy.
3. Significant results
Response to Stress Protocol
Subjective response
As assessed by visual analogue scale (VAS), patients with UC and healthy volunteers (HV) rated the stress protocol as stressful with a median increase of 2.5 units in UC (3(1-5) vs. 5.5(4-8), p<0.0001) and 2.5 units in HV (2(2-3) vs. 4.5 (3.5-6.0), p=0.008).
Autonomic response
The mean pulse rate during the stress protocol was increased compared to pre-test values by 7 bpm (median) in patients with quiescent UC (p<0.0001) and by 11bpm in HV (p=0.02). Thirty minutes after the end of the stress protocol, the mean pulse rate had returned to baseline in both groups.
Mean systolic blood pressure increased during the stress protocol by a median of 12 mmHg in patients with UC (p<0.0001) and 9 mmHg in HV (p=0.03). In patients with UC, mean diastolic BP increased by 7mmHg during stress (p<0.0001).
Serum cytokine concentrations
Serum IL-6 and IL-13 concentrations were unaffected by stress in either UC or in HV.
LPS-stimulated cytokine production
The production of TNF-alpha by LPS-stimulated whole blood was increased after stress by 54% (median) in patients with UC (p=0.004) and by 94% in HV (p=0.03). Production of IL-6 rose after the stress protocol by 11% in patients with UC (p=0.04).
Leucocyte Count
Total white cell count (WBC) increased compared to baseline by 16% (median) in patients with UC (p=0.01) and 17% in HV (p=0.04).
Natural Killer (NK) cell numbers
In patients with quiescent UC, median NK cell numbers, expressed as a percentage of lymphocytes and monocytes, were increased by 18% immediately after the stress protocol (p=0.0008) but returned to base line 30 minutes later.
Platelet Activation
The median percentage of platelets expressing p-selectin was increased by 65% in patients with UC (p<0.0001) and by 61% in HV (p=0.04) after the stress protocol compared to the pre-test sample. This percentage fell in the sample taken 30 minutes later but remained elevated compared to baseline (29% in patients with UC (p=0.001) and 22% in HV (p=0.001)).
Platelet-leucocyte aggregate (PLA) formation
The median percentage of leucocytes forming PLAs was increased by 25% in patients with quiescent UC (p=0.004) and by 6% in HV (p=0.03) immediately after the stress protocol compared to base-line. In patients with quiescent UC, PLA formation remained elevated by 25% 30 minutes later (p=0.002).
Rectal peri-mucosal fluid cytokine levels
In patients with UC, the median TNF-alpha concentration in rectal peri-mucosal fluid was increased by 102% after the stress protocol (p=0.03); the concentrations of IL-13, histamine and substance P in the peri-mucosal fluid did not change.
Mucosal reactive oxygen metabolite (ROM) production
The stress protocol increased median mucosal production of ROMs by 475% (p=0.001).
Rectal Mucosal Blood Flow
Rectal mucosal blood flow fell by 22% (median) after stress (p=0.05).
Histological Assessment
17 pairs of pre- and post-stress biopsies were available for assessment. There was no overall change in histological score (24), with 12 pairs scoring zero both before and after the stress protocol. However, in all five patients where there was a degree of inflammation present in the pre-stress biopsy, there was an increase in histological score in the post-stress sample.
Comparison of patients with quiescent UC and healthy volunteers
There were no differences between baseline pre-test values in any of the variables measured between patients with quiescent UC and HV. There were also no differences in the changes elicited by the stress protocol in any variables between patients with inactive UC and HV.
Psychometric questionnaires
Patients with inactive UC scored higher than healthy volunteers on the Hospital Anxiety Depression anxiety scale (HADS-A) (7(5-10) vs. 4(3-8), P=0.02), Hospital Anxiety Depression depression scale (HADS-D) (3(2-4) vs. 0(0-1), P=0.02), Perceived Stress Questionnaire (PSQ) (62(53-70) vs. 50(40-61), P=0.03) and Bradford Somatic Inventory (BSI) (9 (5-15) vs. 4(2-8), P=0.03). Scores were similar for patients with UC and healthy volunteers on the State Trait Anxiety Inventory State scale (STAI-S) (33 (25-40) vs. 31(25-39)), a measure of anxiety at that moment, and State Trait Anxiety Inventory Trait scale (STAI-T) (36(33-43) vs. 32(28-44)), a measure of anxiety over the long term.
None of the measures of long term stress (STAI-T, HADS, PSQ or BSI) correlated with the changes in any of the variables measured in response to the stress protocol. However, scores on the STAI-S scale did correlate with changes in NK cells levels (R=+0.56, p=0.004) and PLA formation (R=+0.43, p=0.01). There were no differences in scores of any of the psychometric questionnaires between patients with UC undergoing the stress or control protocols.
Response to Hypnotherapy Protocol
Data to be analyzed.
Response to Control Protocol
The control protocol caused no changes in any of the variables assessed in either patients with inactive UC or HV.
Lay Summary
Although many patients and doctors anecdotally report that acute psychological stress can precipitate relapse it is not known how this might occur. It is likely that the effects are mediated via interactions with nerves and inflammatory cells in the gut wall. However to date there have been no studies of this relationship in man, nor of the potential therapeutic value of antagonizing the effects of mental stress in IBD patients.
Conventional medical treatment of IBD is not always effective, and is commonly associated with side effects. Many patients ultimately require surgical resection of involved gut. A significant percentage of patients with IBD therefore use alternative, complementary therapy. Hypnotherapy has proven benefits in irritable bowel syndrome (IBS) and non-ulcer dyspepsia and is anecdotally reported to have beneficial effects in IBD. However little work has been done to confirm such claims or to discover how they might be mediated.
The aim of this study was to investigate the hypothesis that inflammation in IBD could be modified by stress levels. We proposed that increased psychological stress worsened inflammation and that relaxation achieved by hypnosis reduced inflammation. We investigated the contrasting pro- and anti- inflammatory effects of an acute psychological stress test and hypnosis on the general and rectal inflammatory response in patients with UC and healthy controls. Before and after these procedures, the level of inflammation in blood samples was assessed by measurement of serum inflammatory mediators and their production by stimulated blood, white blood cell count and platelet function. The function of the rectal lining, where inflammation is most marked in patients with UC, was evaluated by measurement of its release of inflammatory molecules, and by recording of blood flow using a special laser probe. The cardiovascular to stress and hypnotherapy was quantified by measurement of pulse and blood pressure. Each patient’s background level of chronic stress was also measured using various questionnaires.
We found that stress increased the production of inflammatory mediators by stimulated blood, increased total white cell count and activated platelets. In the rectum, the release of TNF-alpha and oxygen free radicals, key inflammatory molecules, was increased. The data relating to hypnotherapy is being analyzed.
We hope that these studies of the effects of psychological stress and hypnotherapy on inflammation in the body as a whole, and rectum in particular, have improved our understanding of the causation of IBD and may lead ultimately to new therapeutic approaches. Indeed, in the longer term, we hope this work may lead to clinical trials of the efficacy and safety of hypnotherapy in patients with UC and Crohn’s disease.